1. “Near-far” self-classification capabilities of EDMG STAs
Authors:
Name
Affiliation
Address
Phone
Alexander Maltsev
Intel
Turgeneva 30, Nizhny Novgorod, , Russia
+7 (831)
Ilya Bolotin
Andrey Pudeyev
Artyom Lomayev
Oren Kedem
Hiroyuki Motozuka
Panasonic
Gaius Wee
Takenori Sakamoto
Alexander Maltsev, Intel

2. Problem statement
In accordance with Candidate Draft P802.11ay_D0.3 in beamforming procedure for the asymmetric link each STA tries to perform negotiation in A-BFT and then, if it fails, STA performs negotiation in Beamforming training (BFT) allocation during the DTI. Thus the far away EDMG STAs in new outdoor dense deployment scenario will perform a lot of useless transmissions in A-BFT producing undesirable collisions and interference.
Therefore, it is desirable to sort the STAs by its range and capabilities, and separate the EDMG STAs that can be served with the legacy procedure (negotiation in A-BFT and access in legacy allocations) from the EDMG STAs that can be served only by using the directional transmission-reception.
Alexander Maltsev, Intel

3. Near-far self- classification (1/3)
The AP should provide its capabilities to the STAs. With knowledge of the AP parameters, its own parameters and RSSI measurements, the EDMG STA can decide, whether its transmitted signal can be reliably received by the AP in quasi-omni mode (in A-BFT and legacy allocations), or it should try to negotiate during the BFT allocation.
The EDMG STA can go through the legacy procedure (negotiation in A-BFT and access in legacy allocations) if the inequality (2) below is satisfied:
(1)
(2)
Alexander Maltsev, Intel

4. Near-far self- classification (2/3)
Substituting (1) in (2) we obtain
We can denote PAPsensitivity as follows: PAPsensitivity = Pmin_sensitivity - GAPadd_gain
where Pmin_sensitivity is the receiver sensitivity in 11ad, and GAPadd_gain is additional sensitivity gain at the receiver of the AP (due to specific implementation)
The final condition to classify STA as “near” looks like:
(3)
Alexander Maltsev, Intel

5. Near-far self- classification (3/3)
To implement this solution in the standard, the DMG Beacon frame, broadcasted by AP in BTI, should contain the quantized value for PCP/AP Coverage Parameter:
PCP/AP Coverage Parameter = (EIRPAP – GAPomni_rx – GAPadd_gain) (4)
3 bits of the SSW field in the DMG Beacon (see Figure 47, in Candidate Draft P802.11ay_D0.32) may be able to contain the quantized value (4):
Codes 0-7: 14 – 42 dBm (4 dB quantization step) B0
B1 B9
B10 B15
B B17
B18
B B22
B23
Direction
CDOWN
Sector ID
DMG Antenna ID
Quasi-omni TX
PCP/AP Coverage Parameter
Reserved
Bits 1 9 6 2 3
NOTE: as far as antenna gain values in (4) may vary depending on the direction, the value of PCP/AP Coverage Parameter may be sector-specific
Alexander Maltsev, Intel

6. The behavior of EDMG STA with “near-far” self- classification capability
Alexander Maltsev, Intel

7. SP #1
Do you agree to insert the following in the Candidate Draft P802.11ay:
“The EDMG STA should support the “Near-far” self-classification capability described in Slides 3-6”
Intel Corporation

8. SP #2
Do you agree to insert the following in the Candidate Draft P802.11ay:
“Bits B19-B21 of the Sector Sweep (SSW) field, transmitted in a DMG Beacon frame should contain the quantized value for the PCP/AP Coverage Parameter, defined by Eq.(4) on Slide 5. The range of possible values of this parameter is 14 – 42 dBm (4 dB quantization step)”
Intel Corporation

9. Motion #1
Do you agree to include the text for “Near-Far” self-classification capabilities of EDMG STAs proposed in ( ay-text-near-far-self-classification-capabilities-of-edmg-stas) to the spec draft?
Alexander Maltsev, Intel

10. BACKUP
Alexander Maltsev, Intel

11. AP Coverage Parameter range estimation
AP Coverage Parameter = (EIRPAP – GAPomni_rx – GAPadd_gain)
EIRPAP – GAPomni_rx
Upper limit: FCC requirements for 60GHz - EIRP < 43dBm. GAPomni_rx > 0
Lower limit: the 11ad standard defines the maximal difference between peaks of directional and quasi-omni patterns as 15 dB ( of 11mc). We are confident that for 11ay, where the large antenna arrays will be used, this difference will be higher:
GAP – GAPomni_rx > 15dB => EIRPAP – GAPomni_rx > 20dBm
GAPadd_gain
Upper limit: this parameter is the gain over the sensitivity defined in Table The values in that table were obtained under assumption of 5 dB implementation loss and 10 dB noise factor. So, the theoretical maximum of GAPadd_gain is 15dB. However, we are confident, that, in reality, this value will not be more than 6dB. GAPadd_gain < 6dB
Lower limit: the value is definitely positive as otherwise it will not satisfy the minimal sensitivity defined in Table GAPadd_gain > 0dB
14dBm < (EIRPAP – GAPomni_rx – GAPadd_gain) < 42dBm
Alexander Maltsev, Intel

12. AP antenna gain value mismatch impact
GAPtx considered in this proposal is a peak AP antenna gain
There will be cases where the signal was received by STA not in the peak gain. So, the value, provided by AP (GAPmax), will be higher than actual value (GAPtx = GAPmax – GAP_mismatch).
The AP antenna beamforming mismatch leads to the overestimation of Pathloss in equation (1) and, therefore, can be considered as an additional margin for criterion (3). So, STA will define itself farther than it actually is. The reduction of this effect may be done by proper correction of AP Coverage Parameter.
Alexander Maltsev, Intel